Ananth Tiru

Disrupt your OI strategy

Blog Post created by Ananth Tiru Employee on May 26, 2018

Operational intelligence, OI for short, is not something new. We use it in our everyday lives, factories use it, financial institution uses it, pretty much every vertical uses it.

 

So, what is OI – OI enables making decision in the NOW. NOW refers to a reasonable time in which making decisions yields value for the decision maker. For example, a favorite application of mine, Waze, enables me to make decisions when driving to a destination.  A stock trader realizes value when a trade (buy or sell) is made in an appropriate time window. A better production yield can be realized if a manufacturing floor manager can plan an alternative production plan if an anomaly in productions lines can be detected within a reasonable time window.  The list can go on.

OI typically OI combines the past, present and predicted intelligence to respond in the NOW. Typically, OI is implemented an extension to BI systems, Search and Discovery systems (e.g Splunk) or Big Data systems (Hadoop). While these systems are really good for what they are originally designed for, they typically can be inefficient and ineffective for OI.

Some of the key reasons these systems are inefficient are

  • Typically, they require data to be stored and indexed before applying the rules
  • The rules are typically executed at some scheduled intervals
  • The rules are typically executed on a batch or micro-batch of data. The types of OI computations that can be performed in this mode can be limiting or can get quite complex.
  • Typically, these systems are centralized (e.g. running on the cloud) this would mean the data needs to be hauled, stored, analyzed and then the insights (e.g. an Anomaly detected) needs to be published so it can be consumed (e.g. by an operator) typically at a remote location.

 

After examining various use cases in IT and IoT  and the way OI is implemented, we concluded the implementation of OI needs to be disrupted so OI can be implemented to be faster, cheaper and better. Also, we recognized that while we disrupt OI, we don't rip and replace systems which are designed for gathering other types of intelligence (BI, Discovery ).

The following link provides a good description of our offering and the value it offers. 

Hitachi's solution for OI

Outcomes